kvlak limn

Labor & Workplace · MSHA Part 50

Down the Mine

The factory floor answers to OSHA. The mine does not. Coal, metal, sand, and stone answer to MSHA, the one agency that counts both halves of the only ratio that matters - every injury and death, over the hours actually worked. That denominator is the whole point. Divide by hours and a shrinking industry can no longer pass off closed mines as safe ones, and a quarter century of falling body counts finally has to say which it was.

2.53 reportable injuries per 200,000 hours worked all US mining, 2024
1/18M one death for every 18 million hours underground and above deaths per employee-hour, 2024
560M employee-hours worked per year span 2000-2025 Illustrative

The Rate, State by State

I · Near You

Start with the map the title promises. Each mining state is shaded by its injury rate - reportable injuries for every 200,000 hours worked, the “per 100 full-time miners” basis MSHA reports on. Because it is a rate and not a count, Appalachia’s handful of deep coal states burns darker than the sprawling open-pit West. The national rate sits near 2.5; the top of the ladder runs well above it. States with no material reportable mining are left blank, not zero.

Alabama: 3.2 injuries per 200k hours (Underground coal) Alaska: 2.2 injuries per 200k hours (Zinc & gold) Arizona: 2.3 injuries per 200k hours (Copper (metal)) Colorado: 2.4 injuries per 200k hours (Coal & metal) Florida: 2.3 injuries per 200k hours (Phosphate & stone) Georgia: 2.2 injuries per 200k hours (Stone & clay) Indiana: 2.6 injuries per 200k hours (Surface coal) Kansas: 2.4 injuries per 200k hours (Stone & salt) Maine: no material mining Massachusetts: no material mining Minnesota: 2.5 injuries per 200k hours (Iron ore) New Jersey: 1.9 injuries per 200k hours (Stone & sand) North Carolina: 2.0 injuries per 200k hours (Stone & aggregate) North Dakota: 1.8 injuries per 200k hours (Lignite coal) Oklahoma: 2.5 injuries per 200k hours (Coal & aggregate) Pennsylvania: 3.0 injuries per 200k hours (Anthracite & stone) South Dakota: 2.2 injuries per 200k hours (Gold & stone) Texas: 2.2 injuries per 200k hours (Stone & aggregate) Wyoming: 1.9 injuries per 200k hours (Surface coal) Connecticut: no material mining Missouri: 2.6 injuries per 200k hours (Lead & stone) West Virginia: 3.9 injuries per 200k hours (Underground coal) Illinois: 2.9 injuries per 200k hours (Underground coal) New Mexico: 2.5 injuries per 200k hours (Coal & potash) Arkansas: 2.4 injuries per 200k hours (Stone & bromine) California: 2.0 injuries per 200k hours (Aggregate & cement) Delaware: no material mining District of Columbia: no material mining Hawaii: no material mining Iowa: 2.3 injuries per 200k hours (Limestone) Kentucky: 3.6 injuries per 200k hours (Underground coal) Maryland: 2.6 injuries per 200k hours (Coal & stone) Michigan: 2.4 injuries per 200k hours (Iron & stone) Mississippi: 2.1 injuries per 200k hours (Aggregate) Montana: 2.3 injuries per 200k hours (Coal & metal) New Hampshire: no material mining New York: 2.1 injuries per 200k hours (Stone & salt) Ohio: 2.6 injuries per 200k hours (Coal & stone) Oregon: 2.1 injuries per 200k hours (Aggregate) Tennessee: 2.7 injuries per 200k hours (Coal & stone) Utah: 2.8 injuries per 200k hours (Coal & copper) Virginia: 3.4 injuries per 200k hours (Underground coal) Washington: 2.0 injuries per 200k hours (Aggregate) Wisconsin: 2.2 injuries per 200k hours (Sand & stone) Nebraska: 2.0 injuries per 200k hours (Aggregate) South Carolina: 2.0 injuries per 200k hours (Stone) Idaho: 2.3 injuries per 200k hours (Silver & phosphate) Nevada: 2.1 injuries per 200k hours (Gold (metal)) Vermont: 2.1 injuries per 200k hours (Dimension stone) Louisiana: 2.3 injuries per 200k hours (Salt & aggregate) Rhode Island: no material mining
Shade encodes reportable injuries per 200,000 employee-hours worked (five quantile classes). Alaska and Hawaii are inset by the Albers USA projection; US territories fall outside its frame and appear only in the table.
Highest injury rate
  1. 01 West Virginia 3.9 /200k
  2. 02 Kentucky 3.6 /200k
  3. 03 Virginia 3.4 /200k
  4. 04 Alabama 3.2 /200k
  5. 05 Pennsylvania 3.0 /200k
Every mining state, in a table
State Injuries / 200k hrs Fatalities (span) Hours (M/yr) Dominant work
West Virginia WV 3.9 150 36 Underground coal
Kentucky KY 3.6 112 26 Underground coal
Virginia VA 3.4 52 10 Underground coal
Alabama AL 3.2 47 14 Underground coal
Pennsylvania PA 3.0 60 34 Anthracite & stone
Illinois IL 2.9 26 16 Underground coal
Utah UT 2.8 39 10 Coal & copper
Tennessee TN 2.7 21 9 Coal & stone
Ohio OH 2.6 26 13 Coal & stone
Indiana IN 2.6 19 13 Surface coal
Missouri MO 2.6 26 15 Lead & stone
Maryland MD 2.6 10 5 Coal & stone
New Mexico NM 2.5 21 11 Coal & potash
Minnesota MN 2.5 17 14 Iron ore
Oklahoma OK 2.5 17 9 Coal & aggregate
Colorado CO 2.4 26 12 Coal & metal
Michigan MI 2.4 14 11 Iron & stone
Kansas KS 2.4 14 9 Stone & salt
Arkansas AR 2.4 12 8 Stone & bromine
Arizona AZ 2.3 34 24 Copper (metal)
Florida FL 2.3 26 18 Phosphate & stone
Montana MT 2.3 17 8 Coal & metal
Iowa IA 2.3 12 9 Limestone
Louisiana LA 2.3 14 9 Salt & aggregate
Idaho ID 2.3 12 7 Silver & phosphate
Texas TX 2.2 60 42 Stone & aggregate
Georgia GA 2.2 19 13 Stone & clay
Wisconsin WI 2.2 10 8 Sand & stone
South Dakota SD 2.2 9 5 Gold & stone
Alaska AK 2.2 10 6 Zinc & gold
Nevada NV 2.1 41 30 Gold (metal)
New York NY 2.1 15 12 Stone & salt
Oregon OR 2.1 9 7 Aggregate
Mississippi MS 2.1 9 6 Aggregate
Vermont VT 2.1 5 3 Dimension stone
California CA 2.0 41 32 Aggregate & cement
North Carolina NC 2.0 14 11 Stone & aggregate
Washington WA 2.0 10 8 Aggregate
South Carolina SC 2.0 10 8 Stone
Nebraska NE 2.0 8 6 Aggregate
Wyoming WY 1.9 34 26 Surface coal
New Jersey NJ 1.9 7 6 Stone & sand
North Dakota ND 1.8 9 6 Lignite coal

Injury rate is the reference-year reportable-injury count times 200,000, divided by employee-hours worked in the state. Fatalities are the span total (2000-2025). Illustrative stand-in values - see Methodology and HANDOFF.md for the exact MSHA join that fills them.

Which Work Hurts Fastest

II · The Ledger

Not all mining is equally dangerous, and hours worked is what lets you say so. Rank each kind of work by injuries per 200,000 hours and the order is stark: underground coal, on a fraction of the industry’s hours, tops the ledger; sand and gravel move mountains of material at roughly half its rate. The bars share one scale, so length is the comparison; the vertical tick is the national average of 2.5. Every rung is coal or metal, tagged and labeled.

  1. 01 Underground coal Coal 3.9
    9% of national hours worked · 300 deaths over 2000-2025
  2. 02 Underground metal / nonmetal Metal / nonmetal 3.1
    6% of national hours worked · 90 deaths over 2000-2025
  3. 03 Coal prep plant & mill Coal 2.5
    7% of national hours worked · 45 deaths over 2000-2025
  4. 04 Crushed stone Metal / nonmetal 2.4
    22% of national hours worked · 210 deaths over 2000-2025
  5. 05 Nonmetal (potash, salt, phosphate) Metal / nonmetal 2.3
    8% of national hours worked · 70 deaths over 2000-2025
  6. 06 Surface coal Coal 2.1
    14% of national hours worked · 170 deaths over 2000-2025
  7. 07 Surface metal (open pit) Metal / nonmetal 2.0
    13% of national hours worked · 110 deaths over 2000-2025
  8. 08 Sand & gravel Metal / nonmetal 1.8
    17% of national hours worked · 130 deaths over 2000-2025
  9. 09 Office & other surface Metal / nonmetal 0.6
    4% of national hours worked · 25 deaths over 2000-2025

Danger Against Exposure

III · Rate vs Hours

The ledger ranked work by rate alone. Plot that rate against how much of the industry’s hours each kind of work actually burns, and a second truth appears. Underground coal sits high and far right - lethal per hour, yet a thin slice of the workforce. Sand, gravel, and crushed stone sit low and wide: a gentler rate spread across so many hours that the bubbles - sized by the deaths themselves - stay stubbornly large. Safe-looking work, done at that scale, still fills graves.

0% 5% 10% 15% 20% 25% 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 National avg 2.5 Injuries per 200,000 hours worked → Share of national hours worked → Underground coal: 3.9 per 200k hours, 9% of hours, 300 deaths over 2000-2025 1 Underground metal / nonmetal: 3.1 per 200k hours, 6% of hours, 90 deaths over 2000-2025 2 Coal prep plant & mill: 2.5 per 200k hours, 7% of hours, 45 deaths over 2000-2025 3 Crushed stone: 2.4 per 200k hours, 22% of hours, 210 deaths over 2000-2025 4 Nonmetal (potash, salt, phosphate): 2.3 per 200k hours, 8% of hours, 70 deaths over 2000-2025 5 Surface coal: 2.1 per 200k hours, 14% of hours, 170 deaths over 2000-2025 6 Surface metal (open pit): 2.0 per 200k hours, 13% of hours, 110 deaths over 2000-2025 7 Sand & gravel: 1.8 per 200k hours, 17% of hours, 130 deaths over 2000-2025 8 Office & other surface: 0.6 per 200k hours, 4% of hours, 25 deaths over 2000-2025 9
Horizontal: reportable injuries per 200,000 hours. Vertical: that work’s share of national employee-hours. Bubble area is span fatalities. Bubbles are numbered by injury-rate rank; the key names them. Illustrative values.
  1. 1 Underground coal 3.9
  2. 2 Underground metal / nonmetal 3.1
  3. 3 Coal prep plant & mill 2.5
  4. 4 Crushed stone 2.4
  5. 5 Nonmetal (potash, salt, phosphate) 2.3
  6. 6 Surface coal 2.1
  7. 7 Surface metal (open pit) 2.0
  8. 8 Sand & gravel 1.8
  9. 9 Office & other surface 0.6

The Falling Body Count

IV · Year by Year

Annual mining deaths fell from 85 in 2000 to 31 in 2024, a 64% drop, and the line is real progress. But a raw count moves for two reasons at once: mines got safer, and mines closed. Read this curve, then hold the question - fewer miners, or safer mines? - until section VI settles it. The spikes carry names; the disasters marked here are matters of public record.

0 30 60 90 20002005201020152020 2006 Sago Mine explosion 2007 Crandall Canyon collapse 2010 Upper Big Branch 2000: 85 deaths (47 coal, 38 metal/nonmetal) 2001: 72 deaths (42 coal, 30 metal/nonmetal) 2002: 68 deaths (28 coal, 40 metal/nonmetal) 2003: 56 deaths (30 coal, 26 metal/nonmetal) 2004: 55 deaths (28 coal, 27 metal/nonmetal) 2005: 57 deaths (23 coal, 34 metal/nonmetal) 2006: 72 deaths (47 coal, 25 metal/nonmetal) 2007: 67 deaths (34 coal, 33 metal/nonmetal) 2008: 52 deaths (30 coal, 22 metal/nonmetal) 2009: 34 deaths (18 coal, 16 metal/nonmetal) 2010: 71 deaths (48 coal, 23 metal/nonmetal) 2011: 37 deaths (21 coal, 16 metal/nonmetal) 2012: 36 deaths (20 coal, 16 metal/nonmetal) 2013: 42 deaths (20 coal, 22 metal/nonmetal) 2014: 45 deaths (16 coal, 29 metal/nonmetal) 2015: 29 deaths (12 coal, 17 metal/nonmetal) 2016: 26 deaths (9 coal, 17 metal/nonmetal) 2017: 28 deaths (15 coal, 13 metal/nonmetal) 2018: 27 deaths (12 coal, 15 metal/nonmetal) 2019: 24 deaths (12 coal, 12 metal/nonmetal) 2020: 29 deaths (5 coal, 24 metal/nonmetal) 2021: 37 deaths (11 coal, 26 metal/nonmetal) 2022: 30 deaths (11 coal, 19 metal/nonmetal) 2023: 40 deaths (13 coal, 27 metal/nonmetal) 2024: 31 deaths (9 coal, 22 metal/nonmetal)
Total US mining fatalities per calendar year (coal plus metal/nonmetal). Vertical ticks mark landmark disasters. Illustrative values through 2024; the live MSHA feed runs to 2025 Q2.

2006 Sago Mine explosion - 12 killed. A methane blast trapped 13 miners in West Virginia; only one survived.

2007 Crandall Canyon collapse - 9 killed. A Utah collapse killed six miners; three rescuers died in a second fall.

2010 Upper Big Branch - 29 killed. The deadliest US coal-mine disaster since 1970, in West Virginia.

Every year, in a table
Year Deaths Coal Metal / NM Injury rate Hours (M)
2000 85 47 38 5.40 615
2001 72 42 30 5.20 610
2002 68 28 40 4.95 600
2003 56 30 26 4.70 595
2004 55 28 27 4.45 610
2005 57 23 34 4.20 640
2006 72 47 25 4.05 675
2007 67 34 33 3.85 690
2008 52 30 22 3.70 700
2009 34 18 16 3.45 660
2010 71 48 23 3.35 675
2011 37 21 16 3.20 700
2012 36 20 16 3.05 695
2013 42 20 22 2.95 675
2014 45 16 29 2.85 665
2015 29 12 17 2.75 630
2016 26 9 17 2.70 590
2017 28 15 13 2.68 575
2018 27 12 15 2.64 585
2019 24 12 12 2.62 580
2020 29 5 24 2.58 545
2021 37 11 26 2.60 555
2022 30 11 19 2.56 560
2023 40 13 27 2.55 565
2024 31 9 22 2.53 560

The Coal Share Collapses

V · Who Is Still Dying

Split the falling toll into coal and everything else. In 2000 coal was 55% of mining deaths; by 2024 it was 29%. The coal band does not just shrink with the total - it caves in, as seams close and the workforce leaves. Metal and nonmetal mining, quarries and open pits, holds roughly level. The face of a mining death is changing from a collier to a truck driver at a limestone pit.

0 30 60 90 20002005201020152020 2000: 85 deaths - 47 coal, 38 metal/nonmetal 2001: 72 deaths - 42 coal, 30 metal/nonmetal 2002: 68 deaths - 28 coal, 40 metal/nonmetal 2003: 56 deaths - 30 coal, 26 metal/nonmetal 2004: 55 deaths - 28 coal, 27 metal/nonmetal 2005: 57 deaths - 23 coal, 34 metal/nonmetal 2006: 72 deaths - 47 coal, 25 metal/nonmetal 2007: 67 deaths - 34 coal, 33 metal/nonmetal 2008: 52 deaths - 30 coal, 22 metal/nonmetal 2009: 34 deaths - 18 coal, 16 metal/nonmetal 2010: 71 deaths - 48 coal, 23 metal/nonmetal 2011: 37 deaths - 21 coal, 16 metal/nonmetal 2012: 36 deaths - 20 coal, 16 metal/nonmetal 2013: 42 deaths - 20 coal, 22 metal/nonmetal 2014: 45 deaths - 16 coal, 29 metal/nonmetal 2015: 29 deaths - 12 coal, 17 metal/nonmetal 2016: 26 deaths - 9 coal, 17 metal/nonmetal 2017: 28 deaths - 15 coal, 13 metal/nonmetal 2018: 27 deaths - 12 coal, 15 metal/nonmetal 2019: 24 deaths - 12 coal, 12 metal/nonmetal 2020: 29 deaths - 5 coal, 24 metal/nonmetal 2021: 37 deaths - 11 coal, 26 metal/nonmetal 2022: 30 deaths - 11 coal, 19 metal/nonmetal 2023: 40 deaths - 13 coal, 27 metal/nonmetal 2024: 31 deaths - 9 coal, 22 metal/nonmetal Metal /nonmetal Coal
Annual fatalities stacked: coal on the baseline, metal/nonmetal above. The upper edge is the same total as the previous section’s line. Illustrative values through 2024.
The split, year by year
Year Coal Metal / NM Total Coal share
2000 47 38 85 55%
2001 42 30 72 58%
2002 28 40 68 41%
2003 30 26 56 54%
2004 28 27 55 51%
2005 23 34 57 40%
2006 47 25 72 65%
2007 34 33 67 51%
2008 30 22 52 58%
2009 18 16 34 53%
2010 48 23 71 68%
2011 21 16 37 57%
2012 20 16 36 56%
2013 20 22 42 48%
2014 16 29 45 36%
2015 12 17 29 41%
2016 9 17 26 35%
2017 15 13 28 54%
2018 12 15 27 44%
2019 12 12 24 50%
2020 5 24 29 17%
2021 11 26 37 30%
2022 11 19 30 37%
2023 13 27 40 33%
2024 9 22 31 29%

Fewer Miners, or Safer Mines?

VI · The Denominator

Here is the whole argument in two lines. Set both to 100 in 2000 and watch them part. Hours worked ended the period near 91 - the industry barely shrank. The injury rate fell to 47, less than half. If the falling toll were only a hollowing-out, the lines would travel together; they do not. The daylight between them is safety a raw death count can never show - mines that got measurably less dangerous per hour worked.

0 25 50 75 100 20002005201020152020 2000: hours worked at 100 (of 2000 = 100) 2001: hours worked at 99 (of 2000 = 100) 2002: hours worked at 98 (of 2000 = 100) 2003: hours worked at 97 (of 2000 = 100) 2004: hours worked at 99 (of 2000 = 100) 2005: hours worked at 104 (of 2000 = 100) 2006: hours worked at 110 (of 2000 = 100) 2007: hours worked at 112 (of 2000 = 100) 2008: hours worked at 114 (of 2000 = 100) 2009: hours worked at 107 (of 2000 = 100) 2010: hours worked at 110 (of 2000 = 100) 2011: hours worked at 114 (of 2000 = 100) 2012: hours worked at 113 (of 2000 = 100) 2013: hours worked at 110 (of 2000 = 100) 2014: hours worked at 108 (of 2000 = 100) 2015: hours worked at 102 (of 2000 = 100) 2016: hours worked at 96 (of 2000 = 100) 2017: hours worked at 93 (of 2000 = 100) 2018: hours worked at 95 (of 2000 = 100) 2019: hours worked at 94 (of 2000 = 100) 2020: hours worked at 89 (of 2000 = 100) 2021: hours worked at 90 (of 2000 = 100) 2022: hours worked at 91 (of 2000 = 100) 2023: hours worked at 92 (of 2000 = 100) 2024: hours worked at 91 (of 2000 = 100) 2000: injury rate at 100 (of 2000 = 100) 2001: injury rate at 96 (of 2000 = 100) 2002: injury rate at 92 (of 2000 = 100) 2003: injury rate at 87 (of 2000 = 100) 2004: injury rate at 82 (of 2000 = 100) 2005: injury rate at 78 (of 2000 = 100) 2006: injury rate at 75 (of 2000 = 100) 2007: injury rate at 71 (of 2000 = 100) 2008: injury rate at 69 (of 2000 = 100) 2009: injury rate at 64 (of 2000 = 100) 2010: injury rate at 62 (of 2000 = 100) 2011: injury rate at 59 (of 2000 = 100) 2012: injury rate at 56 (of 2000 = 100) 2013: injury rate at 55 (of 2000 = 100) 2014: injury rate at 53 (of 2000 = 100) 2015: injury rate at 51 (of 2000 = 100) 2016: injury rate at 50 (of 2000 = 100) 2017: injury rate at 50 (of 2000 = 100) 2018: injury rate at 49 (of 2000 = 100) 2019: injury rate at 49 (of 2000 = 100) 2020: injury rate at 48 (of 2000 = 100) 2021: injury rate at 48 (of 2000 = 100) 2022: injury rate at 47 (of 2000 = 100) 2023: injury rate at 47 (of 2000 = 100) 2024: injury rate at 47 (of 2000 = 100) Hours worked 91 Injury rate 47
Both series indexed to 2000 = 100. Hours worked from MinesProdQuarterly; injury rate = reportable injuries per 200,000 hours. Illustrative values.
The index, in a table
Year Hours (M) Hours index Injury rate Rate index
2000 615 100 5.40 100
2001 610 99 5.20 96
2002 600 98 4.95 92
2003 595 97 4.70 87
2004 610 99 4.45 82
2005 640 104 4.20 78
2006 675 110 4.05 75
2007 690 112 3.85 71
2008 700 114 3.70 69
2009 660 107 3.45 64
2010 675 110 3.35 62
2011 700 114 3.20 59
2012 695 113 3.05 56
2013 675 110 2.95 55
2014 665 108 2.85 53
2015 630 102 2.75 51
2016 590 96 2.70 50
2017 575 93 2.68 50
2018 585 95 2.64 49
2019 580 94 2.62 49
2020 545 89 2.58 48
2021 555 90 2.60 48
2022 560 91 2.56 47
2023 565 92 2.55 47
2024 560 91 2.53 47

How Mining Kills

VII · Cause of Death

Ask what a mine disaster looks like and most people picture a roof caving or a methane blast. The classification of every fatality tells a duller, deadlier story: the machines. Powered haulage - trucks, conveyors, mobile equipment - and machinery together account for 39% of deaths, more than every collapse, explosion, and inundation combined. Modern mining kills the way heavy industry kills: someone caught by something moving.

0 50 100 150 200 250 300 Powered haulage Powered haulage: 276 deaths (24% of span fatalities) 276 Machinery Machinery: 173 deaths (15% of span fatalities) 173 Fall of roof or rib coal-linked Fall of roof or rib: 150 deaths (13% of span fatalities) 150 Slip or fall of person Slip or fall of person: 104 deaths (9% of span fatalities) 104 Ignition or explosion of gas or dust coal-linked Ignition or explosion of gas or dust: 92 deaths (8% of span fatalities) 92 Electrical Electrical: 69 deaths (6% of span fatalities) 69 Fall of highwall or face Fall of highwall or face: 63 deaths (5.5% of span fatalities) 63 Handling materials Handling materials: 58 deaths (5% of span fatalities) 58 Falling or sliding material Falling or sliding material: 46 deaths (4% of span fatalities) 46 Other or exposure Other or exposure: 44 deaths (3.8% of span fatalities) 44 Explosives and breaking agents Explosives and breaking agents: 40 deaths (3.5% of span fatalities) 40 Entrapment or inundation coal-linked Entrapment or inundation: 35 deaths (3% of span fatalities) 35
Span-total fatalities by MSHA accident classification, longest bar first. “Coal-linked” marks hazards that fall disproportionately on coal mines. Illustrative values; the live build groups the real CLASSIFICATION field.
Every cause, in a table
Cause (classification) Deaths Share Coal-linked
Powered haulage 276 24% -
Machinery 173 15% -
Fall of roof or rib 150 13% Yes
Slip or fall of person 104 9% -
Ignition or explosion of gas or dust 92 8% Yes
Electrical 69 6% -
Fall of highwall or face 63 5.5% -
Handling materials 58 5% -
Falling or sliding material 46 4% -
Other or exposure 44 3.8% -
Explosives and breaking agents 40 3.5% -
Entrapment or inundation 35 3% Yes

Deaths are span totals (2000-2025). Share is of all mining fatalities over the span. Illustrative stand-in values.

What a Reportable Incident Costs

VIII · The Severity Ladder

Every chart so far has counted an injury as one thing. It is not. Of the roughly 5,391 reportable incidents in 2024, most cost days of work, a few hundred left a permanent disability, and 31 (0.6%) were fatal - the thin dark rung at the far left, the sliver this whole dashboard was built to keep in view. Severity is a ladder, and only the bottom of it makes the news.

Segments under 1% are drawn at a minimum width so they remain visible; the counts below are exact.

  1. Fatal 0.6% 31
  2. Permanent disability 2.2% 120
  3. Days away from work 38% 2,050
  4. Days restricted or transferred 21.9% 1,180
  5. No days lost (medical only) 31.9% 1,720
  6. Occupational illness 5.4% 290

Methodology

X · Notes on the Data

The figures on this page are built to the shape of MSHA Part 50 (Accidents, Quarterly Employment/Production, Mines). MSHA - not OSHA - regulates every US mine under the Federal Mine Safety and Health Act, and mine operators must file a Form 7000-1 for each accident, injury, or illness, plus quarterly reports of hours worked and production. Those two streams are the whole basis of this dashboard: incidents over hours.

Why hours, not headcount

A raw death count falls when mines get safer and also when mines simply close. To separate the two, every rate here divides incidents by employee-hours worked and multiplies by 200,000 - the hours 100 miners work in a year. So “2.53 per 200,000 hours” reads as injuries per 100 full-time miners, comparable across a shrinking industry and between a deep coal seam and an open pit.

What's real, what's a stand-in

Every number on this page is currently Illustrative: representative stand-in figures, shaped to the real MSHA structure, that let the story exist before the bulk ingest runs. They are badged Illustrative in the masthead and never presented as real. The rate, the state map, the sector scatter, the coal-metal split, the cause-of-death ranking, and the severity ladder are all stand-ins in the real column shape. The one exception is the landmark disasters annotated on the trend - Sago (2006), Crandall Canyon (2007), and Upper Big Branch (2010) are matters of public record - though the surrounding yearly totals around them are illustrative. The exact swap to real data - the download, the column schema, and the MINE_ID join - is documented in the repo's HANDOFF.md and src/lib/source.ts.

How each chart is built

The state map is a five-class quantile choropleth on the injury rate. The sector scatter plots that rate against each kind of work’s share of national hours, with bubble area carrying span fatalities. The composition area splits the annual toll into coal and metal/nonmetal from the same yearly rows the trend line draws. The cause ranking rolls fatalities up by MSHA accident classification; “coal-linked” flags a hazard whose deaths fall mostly on coal mines. Every chart reads from one committed derived.json and is rendered to static SVG at build time - there is no client JavaScript on this page.

What you're not seeing

Contractors report separately, and small operators sometimes underreport low-severity injuries, so the true injury count is a floor, not a ceiling. Occupational illness with a long latency (black lung, silicosis) surfaces in the record years or decades after the exposure, so a single year understates the real toll of dust. Hours worked are self-reported by operators. And a state's rate reflects the mix of work done there - a state of open pits will read safer than a state of deep coal even if any single mine is no different.

Compare any two states →


Generated 2026-07-06 00:00 UTC. Source: MSHA Part 50 (Accidents, Quarterly Employment/Production, Mines). Feed: Illustrative stand-in; real feed spans 2000-01-01 to 2025-Q2, refreshed weekly.